Okay I did this thread on another forum and it was cool for a while until certain people killed it off with facts and logic and stuff lol. They missed the point. It's supposed to be fun and show just how much better than computers we are and how those AI guys are far far behind. Not to tell me why it isn't possible.

Anywho here is the original topic



If the brain were a computer...

What would it's computational power be?

In terms of Hard drive size, If we pretend the average life is 60 years (I made that up btw) and say you get 8 hours of sleep a day leaving you 16 hours a day to recieve data thats like 350, 400 hours of memory. Not counting leap years since I'm using pretend lifespan anyway. So roughly 350, 400 hours worth of video and sound files. Also files for data storing touch and smell files. That's the easy bit.

RAM and processor speed. What type of graphics card would you need to work with an input device like the eye? What type of soundcard to work with the eye? What sort of programs would you need to intepret all this data and filter it? But in order to work out how much space that would take we need to know the sampling rate of the input devices, what quality the data is... ie how many pixels and colours does an eye see? Also how much of it gets discarded... can you remember the license plate of every car your eyes saw today or everybodys face you saw last week?

Other questions would be like how much RAM does a brain need to control automatic functions like breathing, circulation, temperature control etc in the background while running programes like dreaming software. How much more is this increased when you're awake and doing many more things at once. Say walking down the street talking on a phone munching a hot dog. That's a lot of stuff happening at once.

Speaking of dreaming software, how much space does a personality take up? So far I've spoken of data input and analysis but not really whatever part of our computer brain is filled up with all the bits that make a person unique.


Finally, how close are modern computers to approaching these limits?

Have fun

Here are two of the people who spoiled my thread elsewhere. Note that I post this here because while it's mostly true,this is missing the point! It's meant to be speculative fun people.

First smartass guy

"The question is essentially unanswerable, first because the brain, although capable of computation, is fundamentally unlike a computer in much of what it does, and second because we lack detailed knowledge about many aspects of brain function.

Just a few points - some from memory, so don't trust me on all of these (after all, I'm not a computer).

1. The "computeristic" (Artificial Intelligence) and "non-computeristic" functionalities of the human mind have been argued interminably. Some AI proponents claim that it is only a matter of time and greater understanding before a computer can be programmed to simulate human cognition. Others, including Roger Penrose in "The Emperor's New Mind" argue that the mind is capable of arriving at answers that even theoretically are non-computable. My own guess is that AI may be theoretically attainable, but we are so far away as to afford little immediate practical relevance to this prospect.

2. Whatever kind of "computer" the mind is, even metaphorically, it does not appear to be a digital one - i.e., it does not process information via on/off switches. Rather, modifying influences of a chemical nature (e.g., neurotransmitters, hormones, etc.) or even a physical nature (heat) can affect information transfer in neural circuits in an almost infinitely graded fashion. In other words, the probability of an impulse transmission between two neurons can vary not merely as being either one or zero, but also as an enormous range of probabilities in between. That's one single circuit, but how many digital circuits would be required to emulate that degree of control? Again, that question is probably unanswerable. There is very likely even a quantum aspect to the variability and unpredictability of neural transmission, which perhaps plays a role in both fallacious reasoning but also creativity.

3. Human "memory" and computer memory are radically different. We do not store all the perceptions or information we have received in some repository within the brain, but rather store clues that permit us to reconstruct memories. As a consequence, we almost never remember anything with exact accuracy, and the accuracy varies enormously according to circumstances. As a computer analogy, consider our memories to be files subjected to enormous and variable degrees of compression. Because of this variation, one cannot specify any level of computational power that would necessarily match the requirements for human memory.

4. There are, I'm sure, other differences between how we and computers think and remember, but I don't think I can remember what they are."



Second Smartass guy

A computer is a calculating machine, and the human brain while capable of calculations, is more of a heuristic/probability engine. To explain, getting a computer to recognize simple objects takes a huge amount of CPU time. Even if one is clever regarding algorithms and such. Regardless of how you do it, the computer will screw up certain things. For example, if you place a red ball next to an apple, the computer will have a difficult time telling them apart, and its success depends very much on things like lighting, where the apples are placed, the angle and so on. An algorithm that could differentiate between the ball and the apple 100% of the time regardless of conditions (within circumstances that a human would be able to) would be extremely complicated and costly.

A human on the other hand, does this kind of thing hundreds of times per second while maintaining a general 'thread of consciousness' for lack of a better term, and also controlling things like motor skills and involuntary things like keeping your heart beating.

So, if I were to hazard a guess, (a very vague guess) I would say that the human mind would have about 70% of the total computing power on the planet, if it were harnessed in order to create a likeness of a human mind..

This is so disappointing.

I thouht that "blind idiot" business was so funny.

umbra

It was

To illustrate just how far behind computers actually are...

http://news.bbc.co.uk/1/hi/technology/6600965.stm

Now this is about an extremely powerful supercomputer. Presumably the best available and designed for this specific task. Yet it manages to simulate half a mouses brain and that for only one actual second.

Now this is about an extremely powerful supercomputer.

Does that mean I could simulate a spider on my 2.8 GHz dual-core Pentium? Spiders are way cooler than mice.

um

Half a spider?

Let me guess:

memory for operating system: 1 billion gig.

Processor speed: 325,000 GHz

Graphics card: Now Im lost...to handle the majority of processing from the 5 senses simulataneously as well as the background operating system of emotions/ cardio/lymphatic/hormonal/ temperature regulation and the rest of the endless list.

I cant even guess.

Half a spider?

No that takes me down to 4 legs.

Might as well do a mouse with that!

um